Vertebral fractures (VF) are the most common fracture among older persons, afflicting 20-35% of women and 15-25% of men >50 yrs, and leading to profound morbidity, increased mortality, and costs exceeding $1 billion in the US annually. There is strong evidence that factors other than low bone mineral density (BMD) influence VF risk. Yet, the biomechanical factors that contribute to fractures are poorly understood. From a mechanical perspective, a VF occurs when loads applied to the vertebrae exceed its strength. Thus, in this competing renewal, we aim to study several factors that our recent work indicates are related to VF risk, either by affecting vertebral loading and/or vertebra strength. We will build upon our prior cross-sectional findings by studying several novel factors that may contribute to incident vertebral fractures in two, well-characterized cohorts: the Framingham Heart Study Multidetector CT Study and the Age, Gene/Environment Susceptibility- Reykjavik Study (AGES). In the last funding period, we showed that estimates of vertebral loading are highly sensitive to several factors, namely variations in spine curvature, trunk muscle size and trunk muscle density that were not accounted for in our (and others') prior studies and likely led to errors in the in vivo estimates of vertebral loading. Our preliminary dat also show that a novel measure of heterogeneity in the distribution of bone density within the vertebral body is associated with vertebral strength in cadaveric specimens and with prevalent VF, even after adjusting for BMD. Thus we propose to conduct a prospective study of incident VF to determine the contribution of 1) spinal curvature, 2) the size and quality of trunk muscles; and 3) the distribution of bone density within the vertebral body to VF. Further, we will test whether a patient-specific factor-of-risk (ie, load-to-strength ratio) that uses state-of-the art approaches to estimate vertebral strength and in vivo spinal loading predicts incident VF better than aBMD or FRAX alone. In summary, this project is significant because it addresses the need to reduce the growing burden of VF. The proposal is highly innovative and efficient by using existing QCT and clinical data from well-characterized prospective population- based cohorts, along with an experienced team and multidisciplinary approach to gain knowledge about the etiology of VF. Completion of the proposed work will ultimately shift the current paradigm for VF etiology beyond BMD, thereby pointing towards new approaches to identify those at risk for fracture and novel interventions to reduce the occurrence of VF.